Do you let your genAI tools lead you to unexpected places?

25 February 2024 | Issue #36

In this issue:

New MIT study reveals designers work better with AI
Adobe lets you chat with Acrobat
Writers will become editors
Google pauses image creation in Gemini (whoops)
I was able to bring something full circle

Researchers at MIT, Harvard, and Toyota (really?) published a paper focused on designers testing out a new generative AI tool named DesignAID. Much like any other image generation tool, DesignAID enables users to type a text prompt to generate imagery.

The users found it "more inspirational, enjoyable, and useful" than searching for images using traditional methods. This outcome isn't surprising. Searching for images can be an arduous task, especially at the end of the day.

The researchers noted that the benefit of image generation is a volume play. It's very easy to generate many ideas and discard the ones you don't like. There's no guilt associated with that because there wasn't any real time or effort invested.

To me, this is another example of devaluing design. Making these things disposable is going to lead to more people failing to see the value that design can provide, as we discussed last week.

Returning to the study, the interesting part was that the researchers created two modes of DesignAID: Direct Input mode and New Idea mode. The Direct Input mode functions similarly to your typical Dall-E or Midjourney. Type what you want, and it outputs some options. The New Idea mode assists users who don't have a clear idea in mind. It allows designers to find inspiration that would help lead them to the image they're looking for.

This New Idea mode also interested people without design backgrounds, which makes sense for the "I'll know it when I see it" type of people.

I've worked with individuals who thrive in this space. I would throw a bunch of ideas against the wall, and then we'd pick a couple of directions to follow. This environment is where I excel because it's 100% creative and then becomes completely collaborative. Once that trust is established, it allows the creative process to quickly progress without getting bogged down with fine-tuning everything to be pixel perfect.

Many of the topics of the study overview are obvious to anyone who has paid any attention to AI image generation. But for me, it helped crystallize how some people prefer to work. I work both ways. For the image above, I was very vague with my prompt because I wanted to see what Midjourney would deliver.

At the end of the day, we still need a human to select which option is the best.

NEWS AT THE INTERSECTION OF MARKETING, DESIGN, & AI

🗎 Adobe Acrobat adds generative AI to ‘easily chat with documents’

Adobe's latest update to Acrobat introduces a generative AI assistant, aimed at revolutionizing how we interact with PDFs. For graphic designers, this could mean a significant shift in how documents are managed and information is synthesized.

The AI Assistant in Acrobat acts as a conversational engine, capable of summarizing documents, answering queries, and even suggesting content, making it easier for designers to navigate through extensive project briefs or research materials.
This feature is particularly relevant for graphic designers who deal with sensitive client information and need quick access to specific details within large files.

Keep reading

✍🏻 AI Copywriting: Why Writing Jobs Are Safe

This article took an interesting approach to AI taking jobs. The headline is a bit of a misdirect because the author goes on to say that writers will no longer write. They'll become editors.

AI won't replace people who are really good at their jobs
It's not there yet, but these systems will be able to write as good as the best writer in the future

Keep reading this author's last article (just kidding)

😥 Google making changes after Gemini AI portrayed people of color inaccurately

Google's AI tool Gemini recently faced criticism for generating historically inaccurate images, prompting the tech giant to temporarily limit its ability to create images of people. This move is particularly relevant for graphic designers who rely on AI for creating diverse and accurate visual content.

Gemini wrongly depicted people of color in historically inaccurate roles, highlighting the need for AI to produce accurate cultural representations.
Google's adjustments aim to address AI's racial bias, reminding designers to use AI tools thoughtfully to avoid misrepresentation.

NEW RESOURCES FOR YOU

Bringing things full circle

This past week I was able to chat with a young designer who is the daughter of one of my all-time favorite teachers from high school.

It was a fun full-circle moment because I got to give back, albeit indirectly, to someone who helped shape who I am.

You can read more about the interaction, and the advice I gave this young designer, on my recent LinkedIn post.

(Also, if I was actively hiring an enthusiastic designer, I would hire Hannah Rawitscher before she's off the market).

Affiliate of the Week

The Content Entrepreneur Expo is coming up in May. I'm looking forward to attending for the third year. I promise you will walk away inspired to try new things. This newsletter and the mission of After Design is a direct result of having attended last year.

Will you find your next great idea at CEX? I hope so.

Use promo code "DESIGN100" to get $100 of admission.